I gave a short talk at IHES in December (and a rather longer one in Marseille, too) on the topic of modifying quantum gravity models so that they contain realistic matter. A lot of work on quantum gravity is done without any matter fields and one gets the impression that matter fields are an optional extra which just make the system more complicated. The icing on the cake, as Chris Isham used to say about topology.
In my talk I suggested that, on the contrary, quantum gravity models with matter can actually be rather simpler than models without matter. This is because the Einstein action is induced by the matter fields, so removing the requirement to put the Einstein action into the theory from the beginning.
Some slides from my talks at Bayrischzell and Oxford are available. I am writing a short paper expanding this.
Update (Feb ’11) The paper is out now. The key observation is that the bosonic part of the gravity+standard model action can be obtained from a functional integral over fermions. It seems a strange idea that physics is really like this, but then how else to account for such a wonderful accident?
[Note: The paper is somewhat preliminary and I hope to get back to it and add some more detail in a revised version later.]
Gravity matters 